Manual Targeting Problems

[CapnScurvy]

A couple of weeks ago I found a new problem with manual targeting and stated I would report on it further when I learned more. The problem has been around from day one (frankly, I've seen this problem before but I never investigated it further) which has to do with the centering point of the subs Stadimeter optical view is "off centered", creating an inconsistent reading when comparing a targets distance at the bow to the stern. This coupled along with inaccurate mast heights, and inaccurate optical field of view sizes, has plagued the game in preventing the use of manual targeting in anything but close "can't miss" firing solutions. Of course, if you use the automatic targeting option this problem does not exist. However, I would hope everyone tries manual targeting since this "simulation" offers much more than a point and shoot game.

To start, I placed one of my favorite ships to pick on (the Hiryu CV) circling a sub at 8 stationary points, at about 914 meters (1000 yards) distance. Using the stock games mast height of 31 meters I checked each targets Stadimeter found range and the following was determined.





The Balao sub was facing South with the bow reading of 758 meters found with the stadimeter.





The reason why the reading is 758 meters and not the true distance of 921 meters is the fact that the Hiryu mast height should be 37.4 meters tall not 31 meters as the stock game has it. So, you get thrown a curve with inaccurate mast heights (in this case off by 163 meters) and as the comparison to the stern target shows another 12 meters difference.

I tried another target. This time a Northampton CA set up the same way.




Again, the reason why the Stadimeter's returns are not near the true distance is because the Northampton's stock mast height is listed as 47 meters. It should be reading 49 meters.




In both tests, the inconsistency between the bow and stern views proved themselves. What it means is that roughly a half meter in mast height is the difference between the bow reading and the stern reading (not to mention the basic height inaccuracy found in these two ships).

So, how to correct this without causing too much of a problem with other factors? That's what I've been looking at and have had quite a time over these last couple of weeks.

The problem is, one fix won't fit all the subs!! I tried the one fix idea with simply changing the Cameras.dat file and this is what I got.







Yep, it's like an "out of body experience". Your looking at the back side of the Attack Periscope that is being used to view the target?! I could "clip" the scope out of the picture but this created other issues that just wouldn't look right.

Not to mention I found this "fix" didn't work with every sub. Yes, that's right. Each sub conning tower is positioned differently. So to make a fix, each tower needed to be checked (there are 18 of them) and each needed to have the tower moved the appropriate amount. I did this through the sub model, trying different factors, and finally finding the right fit for each sub. I will admit, the S Class boats are correct as they stand. No need to change the positioning of the conning tower for them. The others were not quite the same. The Balao and Porpoise were the worst, Gato and Tambor almost as much.

A couple of problems arose.




So, the deck guns had to be moved as well.

The worst problem is with the Balao rear antenna post which has been placed into the center of the torpedo refit hatch.




But, this is only a cosmetic issue which does not effect anything else.

With the correction of the stadimeter centered point and the rechecking of the height reference point for each ship, I should have a completed fix for the Optical Targeting Correction soon.

[timmyg00]

Will SCAF fix that? I’m running RSRDC for stock 1.5 (no mega-mods), 3000yd bearing Tool, 3D Radar/TDC, and Show Air Contacts. Is there a SCAF version for me? I do manual targeting all the time (usually with the Dick O’Kane method) and I am making out pretty well, but I’m sure I could benefit from some optical correction!

Thanks

TG

[Daniel Prates]

Great research, capt' scurvy.

It should be pointed out, though, that no perfectly precise measurement could be possible anyway. Ships are constantly going up and down in the water, and weight variations makes the ship float with more or less immersion. The mast height is never a 'definitive' measurement. Measuring how much weight a ship is carrying is done by a method called 'draft survey', which measures how immersed is the ship, before and after being unloaded. And I can tell you, its a matter of several meters, or say, 10 feet!

We're dealing with cargo ships, are we not? How do we know if the target is loaded or unloaded? Because the distance of the top of the mast (or any other part of the ship for that matter) will be closer to the waterline when loaded, and farther when unloaded.

So having such thing as a perfect stadimeter reading was impossible (as it still is), even today. Unless you try to guess if the ship is coming or going from the embarkment port. If you knew it was ferrying supplies, or bringing in raw materials, i thing it would be wise to take the stadimeter and input a mast heigh which is several meters lower.

The stadimeter works with the idea of parallax, which, for those who don't know already, is this: your eyes move inwards to see close objects, and outwards to see farther objects. If you know the distance between your pupils, and the angle that is made by the two lines of sight (one for each eye), then it is a simple matter of trigonometry to find the range to the viewed object. The stadimeter does exactelly that. If you imput the mast height from the waterline, and split the image in two, you get the parallax angle and, thus, distance.

But being the mast height a variable info, how can you get PERFECT distance readings? how do you know if the ship is heavy or light? If it has burned up all it's fuel already?

In fact, how do we even know if the game takes this into consideration?

[Daniel Prates]

Still on the subject, this is how parallax works towards rangefindind:




Also, the equation needed to calculate parallax errors is this, of which I proudly confess that I understood DIDDLY SQUAT:

[CapnScurvy]

Originally Posted by timmyg00

Will SCAF fix that? I'm running RSRDC for stock 1.5 (no mega-mods), 3000yd bearing Tool, 3D Radar/TDC, and Show Air Contacts. Is there a SCAF version for me?

No, SCAF will not fix this problem.

As a matter of fact I would not use SCAF since it's "height" correction was based on the premise that the subs center point of calculating range was equal and consistent in whatever direction you took a reading. As I point out, this is not the case.

The truth is, some of the ship "heights" in SCAF are accurate because they were based on using the targets position either directly perpendicular to the test sub (usually a Porpoise class). As the two illustrations of the Hiryu or Northampton show, the positions of the targets either at the 90 degree or 270 degree relative bearing are consistent with each other. Basing a corrected height from those positions will give a correct mast height. However, basing a height of a target in the frontal bow position (which I did) will create an error close to a half a meter off. A half a meter height difference at only 1000 yards will create about a 20 yard error in found range with manual targeting. The error increases by several fold as the true target distance increases.

My plan is to correct the Optical Targeting Correction mod (which I consider the replacement for SCAF) and continue to make specific versions for other mods.


==========================

Originally Posted by Daniel Prates

We're dealing with cargo ships, are we not? How do we know if the target is loaded or unloaded? Because the distance of the top of the mast (or any other part of the ship for that matter) will be closer to the waterline when loaded, and farther when unloaded.

Good consideration, but the game does not take into account this fact. The game makes no distinction between a loaded ship or not. The ships are constantly positioned at the same height.

You mention the stadimeter is not perfect, and I do not disagree. The game sets several factors into play regardless of whether the "height" measurements are correct or not. One of these is the fact that each separate pixel line will give a different stadimeter found range when compared to its "adjacent" pixel line. This difference between adjacent pixel lines will be less toward the upper sections of the scope, and greater toward the waterline. This factor is based on the idea that a target at a greater distance (closer to the horizon) will give a greater amount of inaccuracy when compared to a closer target (which fills the scope view, having its mast top toward the top of the view). The point is, each pixel line between the scope's horizon (waterline) and top of view, will give a different found range when compared to its adjacent pixel line. For a typical target at a 1200 yard distance the difference between pixel lines are about 8 yards each. Throw in lighting conditions, rough seas, an unsteady hand clicking the mouse, and you aren't going to get an accurate range to target even with a correct mast height.

I just don't think the mast heights should be off as they are. Any real Captain who missed a target due to an incorrect mast height would have "red penciled" his Recognition Manual with what he thought was the more accurate measurement in the event of running into the same target again. I'm simply giving a player the accuracy he "should" have when dealing with manual targeting.

By the way, for you TMO 2.0 players. The mod has the Hiryu mast height at 20 meters. Any guesses as to what that figure will give you at the same 1000 yard true distance as in the above illustration? Ever hear of not being able to hit a bull in the ass with a bass fiddle?

[I'm goin' down]

and I thought I was just a bad shot!

[timmyg00]

It looks like Optical Targeting Correction for RSRDC v550 will be what i need (I just have to verify that my version of RSRDC is correct at v550...)

The description mentions the following:

Additionally, an in-game Omnimeter and Range Dial tool were designed to aid in using the new optical views, and to allow a manual range input to the TDC. This in turn, provided an additional desire to have the radar units work as intended to help in determining range to a found target. Rework of the Radar units display to provide an accurate range and bearing is accomplished by an “overlay” for the screens. The stock radars did not show objects at viewable distances, let alone further distant objects. The radar parameters have been changed to allow for them to work as expected.

Is this rework compatible with the 3D TDC/Radar Range Unit Mod?

Thanks again,

TG

[CapnScurvy]

Timmyg00, yes the "OTC for RSRDC v550" will be the version you would use with your set-up.

Please hold off on downloading it though, while I recheck and change its some 150 ships for correct height (I believe RSRDC adds about 45 new ships to stock). I'm working on adding the centering Stadimeter fix to all the OTC mod versions I've released and correct a found problem with a couple of the "optional mods" I have bundled with OTC that cause a CTD when used (these are the two "Realistic Scopes" optional mods). I'm also going to release two new versions that will be compatible with the stock SH4 game patched to only 1.4 (which a lot of players still use), and RFB 1.4.

As an extra "optional mod" I'm adding a much harder Japanese AI to the game. There will be more planes, with greater capacity to hunt you down. The Japanese warships will have an higher level of "awareness" to find you and prevent your presence from becoming a problem. Their depth charges will be more deadly, and more prolonged. A few other changes that will keep you on your guard when making an attack will be added. All in all, a much harder opponent. I've named the optional mod "Tokko's Revenge" (Tokko meaning "special attack" or "tactics", which also was referred to as Kamikaze). We will see if you like it?

As far as OTC working with the mods you mention? I don't know. I've never tried to put the two together to see what's what. Off the cuff, I would say they do not work together.